<p>The pathophysiology of how a maternal infection induces fetal inflammation and subsequently premature birth is a growing area of research. The <em>ex-vivo</em> dual closed-loop placental perfusion model has been widely used to study placental physiology. To address the association between bacterial chorioamnionitis and fetal inflammation, TNF-α induction following lipopolysaccharide (LPS) challenge – a pyrogen of Gram-negative origin – was measured in the perfusion model. Preliminary analysis of perfusates unexpectedly revealed markedly elevated levels of TNF-α in control and LPS-treated groups indicating contamination of material(s) capable of activating innate immune responses.</p> <p>To identify source(s) driving high background TNF-α expression in perfusates, bovine serum albumin (BSA) – the chief component of the perfusion media – the perfusion system and the materno-feto-placental unit were independently examined. To validate a cleaning protocol effective in LPS removal, acid-base and oxidative depyrogenation techniques were also additionally assessed in the perfusion system.</p> <p>Using TNF-α as a surrogate marker of contamination, high background TNF-α expression in previously conducted placental perfusions were attributed to (1) LPS contaminated perfusion media and (2) LPS build up in the perfusion system. Additionally, results from depyrogenation experiments revealed both acid-base and oxidative techniques effectively reduced LPS buildup in the perfusion system to levels that were in accordance with FDA guidelines for medical equipment (< 0.5 EU/mL). Thus, to circumvent LPS-derived contamination placentas should be perfused using endotoxin-free perfusion media and the perfusion system should be cleaned with acid-base or oxidative depyrogenation techniques prior to its use.</p> / Master of Science (MSc)